An examination of the literature (see pages 10 to 46) shows that many of the researches have been carried out with short periods ranging from 2 hours to 10 or 15 minutes; the majority of the experiments were made in 15-minute periods. In our study of the metabolism after food a large proportion of the basal metabolism experiments were likewise made with these short periods.

Critique Of The Short-Period Method

This method is particularly adapted for experiments with a respiration apparatus with which the gaseous metabolism may be determined and the heat output computed from the results. Such experiments are carried out with considerably less expense and the use of intricate calorimetric apparatus is avoided. Furthermore, comparable values for the metabolism may be obtained on the same day; thus one may be certain of a specially determined and reliable base-line each day, with accurate determinations of body-temperature, pulse rate, and respiration rate.

With both the 24-hour period and the 8-hour period, various time-consuming observations must be made and much duplicate data secured before a satisfactory average is obtained. With the short-period base-line, values may be quickly obtained. Indeed, the results of one or two periods may be rejected on account of extraneous muscular activity of some definite nature, and a true base-line may be found from the average of the other periods with more constant results. To avoid possible activity in the rest periods, it has been the excellent custom in Johansson's laboratory in Stockholm to alternate the periods of complete rest with periods of moderate activity and not to insist upon the tedium of an arbitrarily imposed complete muscular rest for any great length of time.

After two or three periods without food have been obtained with closely agreeing results, and the constancy of the base-line established for that day, food may be given the subject and a series of observations made for from 1 to 6 hours, or even longer. With 15-minute periods it is possible to make an observation practically every 30 minutes. The course of the metabolism after the taking of the food can thus be closely followed and a satisfactory curve obtained showing the immediate effect, the maximum increment, and the gradual cessation of the excess heat production. The short-period method is thus more especially fitted for studying small variations in metabolism and particularly the rapidly occurring and disappearing changes.

Although sufficient material is obtained by this method of measurement to draw a graphic picture of the metabolism, the conditions are still not ideal, as the measurements are not continuous and small variations and possible compensation may thus be lost, especially if the intervals between periods are lengthened from any cause.

This method of determining the metabolism for comparison purposes is further open to criticism in that the assumption must be made that the metabolism remains constant throughout the experimental day, whereas the results may be affected by a daily rhythm or variation. The question may be fairly asked: When no food is taken, is the metabolism the same at 3 p. m. as at 9 a. m.? In other words, if a base-line is determined at 9 a. m., food is given at 10 a. m., and the influence of the food is followed until 3 or 4 p. m., can it be assumed that the increment noted at 4 p. m. above the base-line found at 9 a. m. is due exclusively to the influence of food, or is it due in part to a daily rhythm? Johansson found in his experiments, which were carried out with precautions to maintain absolute muscular repose, that the time of day had but little or no influence upon the carbon-dioxide excretion.1 In considering the results of our experiments made by the short-period method, this question of constancy in the basal metabolism from hour to hour may be discussed intelligently, for a large amount of data is available from which conclusions may be drawn.

Discussion Of Results Of Short-Period Experiments

Aside from a few experiments in which the Tissot apparatus was used,2 the universal respiration apparatus3 was employed exclusively for the short-period experiments. The experiments usually began between 8 and 9 a. m. and continued until noon, and sometimes later; the periods as a rule varied but little from 15 minutes in length. In some instances the experiment was 18 periods in length, but the majority were from 5 to 6 periods long.

The data for all of the subjects with whom experiments of five or more periods have been made have been collected and tabulated; the results obtained with over 30 subjects are thus available for comparison. With several subjects the experiments were made at intervals during a period of five or even six years, but with the majority they were made in a period of approximately two months or even two weeks. An abstract of similar data, which includes nearly all of the subjects with whom we are dealing in this publication, has been given in a previous paper from this laboratory.1 Emphasis was there laid upon the variations in the average maximum values for the oxygen consumption, using the average minimum value for a basis. Inasmuch as it is important to note the actual variations which may be observed in a long series of experiments of this kind, the data will be considered in more detail in this publication.

1Johansson, Skand. Arch. f. Physiol., 1898, 8, p. 103. Magnus-Levy likewise states that the time of day has no influence upon the metabolism. (Magnus-Levy, Arch. f. d. ges. Physiol., 1894, 55, p. 32).

2Tissot, Journ. de physiol. et de pathol. gen., 1904, 6, p. 688.

3Benedict, Am. Journ.Physiol., 1909, 24, p. 345; Deutsch. Arch. f. klin. Med., 1912, 107, p. 156. See, also, p. 202 of this monograph.

As it would be impracticable to print all of the material obtained, three typical subjects have been selected and the carbon-dioxide production and oxygen consumption per minute for the individual periods with these subjects have been tabulated. The data for the other subjects are given in abstract. In the tables showing the detailed results, the day has been subdivided into half-hour periods and the data for the individual 15-minute periods of the experiment placed according to the time the observations were made. The values given under "first period" commonly represent those obtained in the experimental periods which occurred between 8 and 8h30m a. m. At the bottom of the tables are given the average values for each 30-minute period throughout the series, thus indicating the average course of the metabolism throughout the day. The minimum and the maximum values for each period are also given and the mean variations of the individual values from the averages. In the extreme right-hand column the averages for the individual experiments are placed, showing the course of the metabolism throughout the months or years of the study.

In studying the results given in these tables, emphasis should be laid only upon the average values and no particular significance given to single values like the maximum and minimum data. This is in accordance with the custom of this laboratory, as such values are liable to technical errors and physiological variations which must necessarily creep into experiments as complicated as these. The average values for the day are drawn from at least two results and usually three or more, while those for consecutive periods are computed from 5 to 42 periods. It should furthermore be remembered that the data for the oxygen consumption give a more logical basis for discussion than those obtained for the carbon-dioxide elimination. This is due to the remarkable influence upon the carbon-dioxide production of the character of the previous diet,2 the large variations in the calorific equivalent of carbon dioxide with varying respiratory quotients, and the possibility of an over-ventilation of the lungs accompanied by an excessive carbon-dioxide production.

1Benedict, Journ. Biol. Chem., 1915, 20, p. 263, table 4. 2Benedict and Higgins, Am. Journ. Physiol., 1912, 30, p. 217.